HandSeg: An Automatically Labeled Dataset for Hand Segmentation from Depth Images
This work addresses the need for better hand segmentation datasets for computer vision researchers, though it is incremental as it builds on existing depth-sensing and annotation techniques.
The authors tackled the problem of creating high-quality annotations for depth-based hand segmentation by developing an automatic method using RGBD sensors and colored gloves, resulting in a large-scale dataset that enables training algorithms to distinguish between two hands where existing datasets fail.
We propose an automatic method for generating high-quality annotations for depth-based hand segmentation, and introduce a large-scale hand segmentation dataset. Existing datasets are typically limited to a single hand. By exploiting the visual cues given by an RGBD sensor and a pair of colored gloves, we automatically generate dense annotations for two hand segmentation. This lowers the cost/complexity of creating high quality datasets, and makes it easy to expand the dataset in the future. We further show that existing datasets, even with data augmentation, are not sufficient to train a hand segmentation algorithm that can distinguish two hands. Source and datasets will be made publicly available.